On 19/03/2024 07:59, Ayush Singh wrote: > On 3/19/24 11:02, Krzysztof Kozlowski wrote: > >> On 16/03/2024 14:06, Ayush Singh wrote: >>> > Are you sure this fits in Linux coding style limit (not checkpatch >>> limit, but the limit expressed by Linux coding style)? >>> >>> >>> Well, I am just using clang-format with column width of 100 instead of >>> 80. The docs now say 80 is prefered rather than mandatory, so well I was >> So you introduce your own style? Then consider it mandatory... >> >>> using 100 since I prefer that. If 80 is necessary or would make review >>> easier than I can just switch to it. >> You do not choose your own coding style. >> >>> >>> I will remove serdev, pwm, clickID and send a new patch with the minimal >>> driver and better commit messages as suggested with Vaishnav. It is >>> important to have good support for mikroBUS boards without clickID as well. >> Best regards, >> Krzysztof >> > > I mean after the whole discussion about 80 vs 100 column line limit a Yeah, and the discussion was saying: use 80, unless code readability is improved by using 100-limit. > few years ago, and change in checkpatch behavior, I thought 100 was an > acceptable column length in the kernel, but I guess was mistaken, and 80 > character is still mandatory? Not sure why there was a change in > checkpatch and docs though. You mistake checkpatch with coding style. What checkpatch tells you, is a suggestion. It's not the coding style. The problem with checkpatch is that people do not understand "why" it proposes something and they implement its warnings literally, thus sometimes decreasing code readability. > > Regardless, I have switched 80 in the next patch since it is mandatory, > and I do not care as long as I can format using a formatter. Please use wrapping as explained in coding style and deviate to 100 character limit only if it increases the readability. Best regards, Krzysztof
